Molecular photochemistry: A general method for localizing conical intersections using the phase-change rule

A photochemical reaction in which a conical intersection is involved is shown to lead to several different products. In particular, thermally allowed products are produced in many cases in addition to photochemically allowed ones. This is a consequence of the electronic wave-function phase-change rule [H. C. Longuet-Higgins, Proc. R. Soc. London Ser. A. 1975, 344, 147], which is a necessary condition for the existence of conical intersections. The rule is used to define the two coordinates along which the conical intersection is formed, and hence its approximate geometry. These two coordinates are defined by the use of the structures of three chemical species on the ground-state surface, termed anchors. Two of the anchors can be chosen as the reactant and the desired product; the third is another possible product. The phase-change rule requires that either one or all the transition states between the three anchors must be phase-inverting, for instance antiaromatic. When only one of the transition states is phase-inverting, a "thermally allowed" product is always involved. The well-known importance of antiaromatic transition states in photochemical pericyclic reactions is explained by their essential role in forming conical intersections. The model provides a rationalization for the properties of many of the recently calculated conical intersections. The phase-change rule provides a simple, chemically oriented method for both the prediction of the course and stereoselectivity of photochemical reactions. It can also be used to reject structures proposed for conical intersections by showing that the conditions necessary for its presence are not fulfilled.